Behavioral Science of AI Adoption
-
95%
of GenAI pilots in companies fail to scale. (MIT)
-
>75 minutes
Course content totaling more than 75 minutes spread across 6 modules and 15 videos. Learn at your own pace!
-
18
Includes 18 downloadable files, worksheets, and templates that allow you to assess AI adoption personally and in your organization.
-
>97%
Over 97% of previous participants in masterclasses, workshops, and executive education recommend the instructor to others.
Why this course exists.
According to MIT's 2025 research on enterprise AI, 95% of generative AI pilots deliver no measurable financial return—not because the tools don't work, but because adoption is fundamentally a behavior-change problem, not a technology problem. The same organizations often see employees quietly using unsanctioned AI tools on their own time, at rates far higher than official usage—proof that people aren't rejecting AI. They're rejecting the specific version of it that was handed to them without any attention to how humans actually change their behavior. This course takes the behavioral science that explains why people resist, over-trust, under-trust, and eventually abandon new tools, and turns it into something practical: a way of seeing adoption clearly, and a small set of skills that make it stick.
What you’ll leave with.
-
A documented case of resistance on your own team, diagnosed against the three psychological forces.
-
A real usage audit comparing what your team says about an AI tool to what the data shows.
-
A Trust Calibration Checklist, tailored to a tool you actually use.
-
One concrete embedding mechanism you've drafted for your own team, with an owner and a start date.
-
A Nudge Audit scoring a real rollout against the EAST framework, with one touchpoint redesigned.
-
A specific if-then plan for turning a stalled intention into an actual behavior.
-
A Personal Adoption Diagnosis and Action Plan, assembled in the course's final video from everything above.
-
-
95% of generative AI pilots deliver no measurable financial return, and it isn't because the technology is weak. This video opens the course with that paradox and the "shadow AI economy" phenomenon, then reframes the whole problem: adoption isn't a rollout event, it's a behavior-change challenge.
-
What looks like stubbornness from the outside is almost always a legible response from the inside. This video introduces the three forces—loss, identity, and uncertainty—that structure the rest of the course, reframing resistance as data rather than defiance.
-
-
-
76% of executives believe their employees are excited about AI; only 31% of employees actually say they are. This video covers social desirability bias and the intention-behavior gap, and teaches you to ask questions that get past both.
-
A login isn't adoption; a shorter workflow is. This video covers what to actually track in usage data—including task substitution versus addition, the one metric most dashboards never show.
-
-
-
Losses loom roughly twice as large as equivalent gains—one of the most replicated findings in behavioral science. This video explains why people defend a clunky process they already own over a better one they don't, and names the four things they're actually protecting.
-
People will bet on known odds over unknown ones, even at identical payoffs. This video connects that classic finding to AI directly: why not knowing why a system gave its answer is scarier than knowing the odds are bad, and how explanation fixes it.
-
The real question isn't "can the AI do this," it's "who am I if it can." This video closes out the three forces with the deepest one—what happens when a tool threatens someone's sense of their own expertise, not just their workflow.
-
-
-
A tool can be right ninety-five percent of the time, and the wrong five percent is the only part anyone remembers. This video covers the landmark research on why people abandon algorithms faster than equally fallible humans, treating one mistake as a verdict instead of a data point.
-
Sometimes the problem isn't too little trust—it's too much, aimed the wrong way. This video covers the research on over-trusting confident-sounding AI output, and why your most experienced team members may be the most exposed to it.
-
Adoption isn't the finish line; calibrated trust is. This video brings the module together into one practical tool—the Trust Calibration Checklist—for keeping trust from swinging too far in either direction.
-
-
-
Employee sentiment toward AI nearly quadruples with strong leadership support, and most managers aren't providing it yet. This video covers the single highest-leverage thing a manager can do for adoption, grounded in research on psychological safety.
-
The average new behavior takes sixty-six days to become automatic; most rollout plans run out of attention in fourteen. This video covers what actually carries a habit past the point where novelty wears off—peer modeling, visible wins, and structural nudges.
-
-
-
A single change to a default moved consent rates from 42% to 82%, with nothing else about the decision changed. This video covers the EAST framework—Easy, Attractive, Social, Timely—for redesigning the environment around an AI rollout instead of relying on a mandate.
-
"I'll use it more" almost never survives a busy week. This video covers implementation intentions—specific if-then plans shown to roughly double follow-through in a landmark meta-analysis—and directly resolves the intention-behavior gap first raised in Module 2.
-
The final video in the course, and the capstone. It introduces no new research—instead, it maps all six modules onto a single four-question sequence for diagnosing any stalled AI rollout, and guides learners through assembling everything they've built into one Personal Adoption Diagnosis and Action Plan.
-
Your instructor.
Eugene Chan, PhD
Eugene (PhD Toronto; MA Chicago; AB Michigan) is a behavioral scientist whose work focuses on the moments when customers are uncertain, skeptical, or under scrutiny, translating insights from consumer psychology into practical strategies that strengthen confidence, reduce friction, and improve adoption across products, services, and communication. His academic work is published in Financial Times Top 50 journals and has been covered by media outlets include Globe and Mail, Men’s Health, and the Wall Street Journal. He has also conducted workshops and masterclasses for leading companies in Canada, Australia, and the United States.
You’ll want to enrol if you want to…
-
... diagnose why a specific AI rollout is stalling, and name the exact psychological force—loss, ambiguity, or identity threat—driving the resistance you're seeing, rather than guessing or defaulting to "change is hard."
-
... read adoption accurately by distinguishing what people say about a tool from what usage data actually shows, and knowing which one to trust.
-
... recognize and correct both failure modes of trust, catching a team that's abandoned a good tool after one visible mistake, and catching a team that's rubber-stamping confident-sounding output without real scrutiny.
-
... lead adoption as a manager using visible modeling and psychological safety to make trying a new tool feel low-risk, and building structural mechanisms—not just training sessions—that survive past the first few weeks.
-
... brief AI tools effectively, supplying the context a colleague would already have, using a simple five-element structure you can apply to any task.
-
... review AI output appropriately, matching your level of scrutiny to the actual stakes of the task, instead of either over-checking everything or rubber-stamping everything.
-
... build a personal practice of improvement, treating every AI interaction as a small opportunity to get sharper rather than a one-off transaction you forget the moment it's over.